Act Like A Chef When Developing Products

product development as experimentation

Topics: Exploratory Product Development | Risk

The goal of product development is to create a “recipe” for producing a product [i], a final result that is sold to customers. The recipe for a tangible product would specify the ingredients, the materials and components to be used. The steps of the recipe use specific amounts of ingredients with specific methods requiring certain labor, tools and machinery to produce components of the product. The components are assembled and further processed into a final product.

In a professional kitchen, experimentation and exploration are integral to creating a new recipe. Sometimes experiments combine ingredients in new ways, use existing methods in new ways, and even create new methods and tools. Sometimes the result is brilliant, sometimes it’s nothing special, sometimes it’s a disaster. But each experiment teaches the chef something new about what works and what doesn’t. The prototype recipe is adjusted and experimentation continues until the final masterpiece is achieved. The recipe can then be used over and over again to produce the same masterpiece.

The Chef Experiments

Like chefs, product developers have to make judgements and choices about ingredients, methods, tools, knowledge and skills to invent recipes for new products [ii]. By definition, each new product uses a different recipe; therefore different judgements and choices have to be made.

What is different? Obviously the desired end result (the new product) is different from existing products. In fact, the target may not even be well-defined at the beginning of the project. Also, the starting point for the project (what is known or not known) will be different. The lack of complete information and definition create uncertainty [iii] in the nature and sequence of activities required to create the recipe.

The specific activities for creating the recipe have to determined contingent on (or based on) what we know and what we need to learn, and that changes during the project. For example, in our pie recipe we want a flakier crust than we’ve been able to create in our other pies. We have the idea that replacing water with vodka [iv] might work, but we need to test that. If it doesn’t work, we’ll need to try something else. Thus the specific activities in developing our recipe are contingent on what we learn in prior activities. The less we know with certainty, the more we need to learn. The more we need to learn, the less certain we can be about specific future activities.

It is hard to predict how these future activities will unfold. Some activities will be familiar to product developers, some will need to be adapted and others will be new. For example, using vodka instead of water in the crust could necessitate changes to subsequent activities. We know a future step in the recipe will be rolling out the dough. What we may discover is that we will need to adapt our method to use more flour, but we won’t learn that until we actually try to roll out the dough. And, we’ll have to experiment to learn the right amount of flour to use.

What we don’t know creates uncertainty about future activities.

Unpredictable future events also create uncertainty. We don’t know with certainty, for example, how the market will change, what competitors will do or what new technologies will arise (no pun intended). We make educated judgments and choices about the recipe before knowing the results. Uncertainty means the outcomes of our judgements and choices may not be what we expect. The result could be positive (the product is more successful than expected), negative (the product is less successful than expected) or neutral (meets expectations).

The problem is: How do we make those judgements and choices in a situation of uncertainty?

One common way to proceed with creating the recipe when there is uncertainty is to use assumptions.  We plan out the process of creating the recipe making assumptions where we lack certain knowledge. Here’s how:

Assumptions act as proxy inputs.

Going back to our pie recipe, here are some possible assumptions:

  • Customers prefer flaky pie crust
  • Substituting another liquid for water will create a flakier crust than our current pie crust recipes
  • Vodka may be a suitable substitute liquid

Based on these assumptions, we believe: this new pie will be successful in the market and we will be successful in creating a flakier crust.

Some assumptions are clearly identified, but often there are tacit assumptions not recognized by the product development team. For example, we may have a team working on the pie crust and a team working on the filling. A tacit assumption may be: The filling will not impact the flakiness of the crust. This is a crucial assumption because if true, the development of the crust and the filling can be pursued in parallel and independently. Doing work concurrently reduces the overall time to deliver the new recipe.

What do we do if the assumptions are wrong?

If any of our assumptions are wrong, our judgements and choices may be faulty. This creates the potential for an unsuccessful product; including items such as going over budget, missing the schedule, not being able to create a product that works and not being able to sell the product. Risk is the potential for an unexpected outcome and is measured on two dimensions: the probability of occurrence and the size of the impact (e.g. money or days lost or gained).

Getting certain assumptions wrong will have a greater impact than getting others wrong. The concept of risk helps us evaluate the impact of uncertainty and incorrect assumptions. We can then target uncertainties, assumptions and risks for resolution through specific activities.

The problem is: How do we identify uncertainties and tacit assumptions?

How do we choose activities to reduce rather than increase uncertainty throughout the product development process?

How do we make assumptions that help, not hinder the product development process?

The product development process guides product developers in determining what activities need to be performed to develop the recipe for a new product. Some processes are prescriptive, meaning they tell developers exactly what activity to do, when to do it, and often exactly how to do it. Prescriptive processes are good for standard and routine processes where inputs are well-defined, uncertainty is low and activities are familiar, like a manufacturing process.

Other processes are descriptive, meaning they help teams think through what needs to be done and the team uses their knowledge and experience to select and execute the appropriate activities. Descriptive processes are best for non-routine purposes, such as product development, where inputs are not well-defined, uncertainty is high and many activities are new.

A good product development process helps the team to identify uncertainties, assumptions and risks, and to choose appropriate activities for creating the product recipe.

Throughout the project, as the team learns new information, the set of uncertainties, assumptions and risks will change. These changes affect their relative importance, impacting which subsequent activities are appropriate. The process has to adapt (modify the activities, scope, budget, schedule) as information is produced to update the uncertainties, assumptions and risks.

The problem is: Many companies still follow a process with prescribed activities that don’t actively identify, measure and reduce risks.

Answer: Exploratory Product Development (ExPD). Learn more about our new process for developing products.

Learn & Adapt: ExPD an Adaptive Product Development Process for Rapid Innovation and Risk Reduction is now available through major booksellers.


[i]  Reinertsen, Donald G., Managing The Design Factory. New York: The Free Press (1997).

[ii] Ian P. McCarthy, Christos Tsinopoulos, Peter Allen, and Christen Rose-Anderssen (2006). New Product Development as a Complex Adaptive System of Decisions, Journal of Product Innovation Management, 23:437-456.

[iii] Situation where the current state of knowledge is such that (1) the order or nature of things is unknown, (2) the consequences, extent, or magnitude of circumstances, conditions, or events is unpredictable, and (3) credible probabilities to possible outcomes cannot be assigned. Although too much uncertainty is undesirable, manageable uncertainty provides the freedom to make creative decisions.

[iv] “Since water bonds with flour to form gluten, too much of it makes a crust tough. But rolling out dry dough is difficult. For a pie dough recipe that baked up tender and flaky and rolled out easily every time, we found a magic ingredient: vodka. Using vodka, which is just 60 percent water, gave us an easy-to-roll crust recipe with less gluten and no alcohol flavor, since the alcohol vaporizes in the oven.”

Comments are closed.

Learn about our other services

For more information, contact:

Mary Drotar, Partner

Strategy 2 Market, Inc.


or complete this contact form

Scroll to Top